How Steep Is This Hill?

weirdly accurate.

    How Steep Is This Hill?

    What is the hill steepness test?

    This hill steepness test puts you in the place of the people in Dennis Proffitt’s 1995 slope study: you judge five hills twice, once in words and once with your hand, and only then see the true slope from the side. Proffitt’s team asked 300 passers-by on the University of Virginia campus about one real hill each, and the result sheet sets your first answer against theirs. You find out how well your eyes read a slope, and whether your words and your hand agree.

    This test puts you in Proffitt’s place on five rendered hills, each a real slope from his 1995 study. On each hill you first say how steep it is by typing a number of degrees, from 0° for flat ground to 90° for a wall, with no line or preview to match by eye. Then you show it with your hand, lifting the end of a small board until it looks as steep as the hill. The camera walks round to the side and shows the true slope, and whether you were spot on (within 3°), too steep or too gentle. At the end you get a steepness factor for your words and one for your hand, your answer as a multiple of the true slope on average, and which of the two was nearer the truth on more hills. One hill, judged before you have seen any answer, is set against the 1995 crowd who judged the same real slope, as a rough percentile.

    It is a perception game, not a vision or fitness test. Your hills are 3D pictures on a screen, with no stereo vision, no walking and a fixed eye height of 1.6 m, while the 1995 numbers come from real hills seen in person. Your board is moved by your finger and seen while you set it, so it is closer to Proffitt’s visual disk than to his unseen palm board. And in the study each person judged one hill without feedback, while here you see every answer, so later hills are not comparable to theirs.

    What the test measures

    One thing, two ways: how steep a hill looks to you in words, and how steep your hand says it is.

    The task

    • Five rendered hills seen from their foot, each a real slope from Proffitt, Bhalla, Gossweiler & Midgett (1995)
    • Say it: type the angle in degrees on a number pad, with no preview line
    • Show it: lift the end of a small board with your finger until it matches the hill
    • Walk round to the side to see the true slope; within 3° counts as spot on

    Your result

    • Your steepness factor in words and with your hand: estimate ÷ true slope, averaged over the hills of 5° and steeper
    • Your average miss for each, and on how many hills your hand beat your words
    • One answer set against the 1995 crowd as a rough percentile, and a chart of your words and hand against their averages
    • Reference slopes such as a wheelchair ramp, the world’s steepest street and a dry sand dune, and your answers as road-sign grades
    • A share card showing only your factor, and a challenge link that carries your first answer in the part after the #; no account

    Before you start

    • About two minutes: five hills, two answers each
    • Works on phones and computers; on a computer you can type the numbers
    • Needs a browser with 3D (WebGL) switched on

    Hill steepness FAQ

    Is this test free? Do I need to sign up to see my result?

    Yes, it’s free, and there’s nothing to sign up for: no account, no email, no payment. Your result appears on this page the moment you finish, and you can share it or play again straight away.

    Do hills look steeper or gentler than they really are?

    That is what the five hills in this test measure, so the page lets you answer before it tells you anything. After each hill the camera walks round to the side and shows the true slope. After the last one, the result sheet sets your words and your hand against the 1995 averages and explains what researchers think is going on, from Proffitt’s own account to Li and Durgin (2010).

    How steep do people think a hill is?

    The result sheet shows it after your five hills, as a chart of the 1995 averages beside your own words and hand. Those averages come from Proffitt and colleagues’ 1995 study of 300 passers-by, each judging one real hill without feedback. One of your hills, judged before you have seen any answer, is set against the people who judged the same slope, as a rough percentile.

    Is your hand better than your words at judging slope?

    Hand against words is the second thing the test measures: on every hill you answer both ways, and the result sheet counts on how many hills your hand beat your words. Afterwards it shows what Proffitt found in 1995 with a palm board tilted by an unseen hand, how he read it, and why Durgin and colleagues (2010) read the same palm-board numbers differently.

    Does a heavy backpack or being tired make hills look steeper?

    The evidence is contested. Bhalla and Proffitt (1999) reported that a heavy backpack or a long run changed people’s spoken estimates of a hill while their palm boards barely moved. But when Durgin and colleagues (2009) gave backpack-wearers a believable cover story, the effect vanished, suggesting people had guessed what the experimenter wanted. The result sheet tells the whole argument once you have finished.

    How steep is the world’s steepest street?

    Baldwin Street in Dunedin, New Zealand, holds the Guinness World Record for the steepest street. The result sheet gives its exact grade and angle after your five hills, next to your own answers and other reference slopes such as a wheelchair ramp and a dry sand dune. Giving the angle here would hand you a yardstick for the hills, so the page saves the number for the end.

    Is the hill steepness test a vision test?

    No. It is a perception game, not a vision or fitness test, and a big over- or underestimate says nothing about your eyesight. The hills are 3D pictures on a screen rather than real hills, and you see every answer as you go, so only your first answer compares fairly with the 1995 study, and only roughly.

    Where the numbers come from

    The averages come from Proffitt et al. (1995), Experiment 1: 300 passers-by on the University of Virginia campus, 30 per hill and 60 on one of them, each judging one real hill; the exact values are printed in Bhalla & Proffitt (1999), Table 2. The percentile treats their answers as a normal curve with SD = SE × √n, because the papers report no standard deviation. Li & Durgin (2010) was read as an abstract. The result titles are the site’s own bands of the words-to-truth ratio, not a norm. The site does not publish its own players’ data.

    How the result is worked out

    Steepness factor

    Your estimate divided by the true slope, averaged geometrically over the hills of 5° and steeper; the gentlest hill is left out because a tiny slope makes the ratio explode. 1× is exact; 2× means you judged hills twice as steep as they are, and 0.5× half as steep.

    Words and hand

    Proffitt’s study measured slope three ways: a spoken number, a disk matched by eye, and a palm board tilted with an unseen hand. This test uses the spoken number and a board you see while you set it.

    Percentile

    A normal curve from the 1995 mean, with the spread derived as standard error × √n because the papers report no standard deviation. A rough guide, not a norm.